deprotonated cysteine: how many water ligands are sufficient to approach solution-phase photooxidation chemistry?

PCCP View Article Online

Published on 17 October 2013. Downloaded by Emory University on 06/06/2014 15:13:44.

PAPER

View Journal | View Issue

Cite this: Phys. Chem. Chem. Phys., 2013, 15, 20496

Oxidation of gas-phase hydrated protonated/ deprotonated cysteine: how many water ligands are sufficient to approach solution-phase photooxidation chemistry? Fangwei Liu, Rifat Emre, Wenchao Lu and Jianbo Liu* We present a study on the reactions of singlet oxygen O2[a1Dg] with hydrated protonated and deprotonated cysteine (Cys) in the gas phase, including measurements of the effects of collision energy (Ecol) and hydration number on reaction cross sections over a center-of-mass Ecol range from 0.05 to 1.0 eV. The aim is to probe how successive addition of water molecules changes the oxidation chemistry of Cys in the gas phase. Hydrated clusters, generated by electrospray ionization, have structures of HSCH2CH(NH3+)CO2H(H2O)1,2 and HSCH2CH(NH2)CO2(H2O)1,2 for protonated and deprotonated forms, respectively. In contrast to 1O2 reactions with dehydrated protonated/deprotonated Cys of which hydroperoxide products all decomposed, reactions with hydrated protonated/deprotonated Cys yielded stable hydroperoxide products, analogous to photooxidation reaction of Cys in solution. We investigated the number of water ligands necessary to produce a stable hydroperoxide, and found that a single water molecule suffices—that is, to relax nascent, energized hydroperoxide in the hydrated cluster by elimination of water. Hydrated protonated Cys shows higher reaction efficiency than the hydrated deprotonated one, particularly with the addition of the second water ligand. Reactions of hydrated protonated/deprotonated Cys are suppressed by Ecol, becoming

Received 4th September 2013, Accepted 16th October 2013 DOI: 10.1039/c3cp53736f

negligible at Ecol Z 0.5 eV. Density functional theory calculations were used to locate reaction coordinates for these systems. Quasi-classical, direct dynamics trajectory simulations were performed for HSCH2CH(NH3+)CO2H(H2O) +

1

O2 at the B3LYP/4-31G(d) level of theory. Analysis of trajectories

highlights the importance of complex mediation in the early stages of the reaction, and illustrates that water can catalyze proton transfer within the hydrated complex.

www.rsc.org/pccp

I. Introduction Singlet molecular oxygen (O2, a1Dg) is produced in biological systems via various paths,1 and oxidation of amino acids by 1O2 is an important process associated with biological aging, diseases,2 and photodynamic therapy for cancer treatment.3 Because of these biological as well as photochemical implications in the atmosphere,4 there has been considerable interest in studying 1 O2-induced amino acid oxidation mechanisms. Most of these experiments were carried out in solution using photosensitization methods5 (referred to as ‘‘photooxidation’’). In photooxidation experiments, 1O2 was generated by sensitizers upon exposure to hu

ultraviolet/visible light,6 i.e., sensitizer ! sensitizer , followed by sensitizer þ3 O2

energy transfer

!

sensitizer þ1 O2 .

Photooxidation

Department of Chemistry and Biochemistry, Queens College and the Graduate Center of the City University of New York, 65-30 Kissena Blvd., Queens, New York 11367, USA. E-mail: [email protected]

20496

Phys. Chem. Chem. Phys., 2013, 15, 20496--20509

of

amino acids often shows complex features with multiple pathways and products depending on pH, oxygen concentration, solvent composition, type of sensitizers, buffer ions, etc. These are partly due to the fact that type I (free radical-mediated) and type II (1O2-mediated) mechanisms might co-exist in photooxidation; as a consequence, other reactive oxygen species besides 1O2 may contribute to oxidation reactions. These factors complicate reaction systems, requiring multiple sets of control experiments in order to obtain a good understanding of oxidation mechanisms in solution. To avoid the complexities/interferences arising from solutionphase experiments and simplify the interpretation of the oxidation mechanisms of amino acids, we have looked at the reactions of 1O2 with protonated and deprotonated amino acids in the gas phase7–11 using electrospray-ionization (ESI)12,13 mass spectrometry and guided-ion-beam scattering methods.14 One advantage of investigating biomolecules in the gas phase is that it allows one to observe single molecules separated from bulk solution environments. In this way, intrinsic reactivity of molecules

This journal is

c

the Owner Societies 2013

View Article Online


Paper

PCCP

can be distinguished from solvent effects. By combining gas-phase experiments, statistical modeling and molecular dynamics simulations, we were able to unravel oxidation mechanisms and dynamics of various gas-phase ‘‘bare’’ amino acid ions, including protonated tyrosine,7 methionine,8 cysteine and tryptophan,9 and deprotonated tyrosine, tryptophan10 and cysteine.11 However, an obvious question for gas-phase experiments of protonated/deprotonated amino acids with 1O2 is how to make these reactions closely resemble photooxidation reactions in biological systems where amino acids are solvated. Because amino acid ions are able to form hydrogen bonds with water molecules, it is often difficult to distinguish the effects caused by the intrinsic properties of amino acid ions from those caused by their interactions with water.15 In theory, this question might be addressed by a comparison of gas- and solution-phase reaction products. In practice, such comparison may become difficult considering the technical limitations of solution-phase photooxidation experiments. In addition, a simple comparison of gas- and solution-phase reactions may not be able to provide information about the effects of individual water molecules on amino acid oxidation. Let us take the reaction of cysteine (Cys) with 1O2 as an example.16–22 Cys is one of the most susceptible residues towards oxidative damage by 1O2 in proteins.1,2,23 Photooxidation of Cys initially forms a persulfoxide intermediate (note that persulfoxide is a well-known key intermediate in reactions of 1 O2 with organic sulfides,24 albeit less common in thiol + 1O2 reactions). Ultimate oxidation products of Cys include cystine, RSO2H, and other species that remain to be elucidated. Photooxidation of Cys shows pH dependence, i.e., in acidic solution, Cys is converted to RSO2H; while in neutral or alkaline media, oxidation is completed by a dark reaction, resulting in cystine rather than RSO2H.25 We have investigated the reactions of 1O2 with protonated and deprotonated Cys, respectively, in the gas phase.9,11 For protonated CysH+, the product channel corresponds to Ca–Cb bond rupture of a hydroperoxide intermediate Cys-SOOH+ accompanied by intra-molecular H atom transfer, and subsequent dissociation to H2NCHCO2H+, CH3SH and 3O2. Deprotonated [Cys-H] has a carboxylate structure HSCH2CH(NH2)CO2, and three product channels were observed for its reaction with 1O2. Dissociation of HSCH2CH(NH2)CO2 to NH2CH2CO2 and CH2S accompanied by quenching of 1O2 has the largest cross section. Two minor channels correspond to the formation of HSCH2C(NH)CO2 + H2O2 and OSCH2CH(NH2)CO2 + OH, respectively. Although oxidation of both protonated and deprotonated Cys yields persulfoxide and hydroperoxide intermediates, none of these was detected in product ion mass spectra. Therefore, oxidation of Cys in the gas phase appears to somewhat deviate from what happens in aqueous solution. To solve this puzzle, in the present paper we report a study on the effects of hydration on oxidation of protonated and deprotonated Cys (as a function of the number of water molecules absorbed). The approach we utilized is to hydrate CysH+ or [Cys-H] by adding water molecules stepwise, and study the reactions of resulting gas-phase hydrated clusters.

This journal is

c


This approach provides detailed information on the reactivity of solvated Cys ions, without interference arising from the presence of bulk water.26 In the following sections, we will describe a guided-ion-beam study together with theoretical simulations of the reactions of 1O2 with gaseous hydrated CysH+(H2O)n and [Cys-H](H2O)n (n = 1–2). As we will show, in contrast to that of CysH+ and [Cys-H], oxidation of CysH+(H2O)1,2 and [Cys-H](H2O)1,2 results in stable peroxide products, similar to solution-phase photooxidation reactions.

II. Experimental and computational details 2.1

Experimental procedures

The experiment was carried out using a home-made guidedion-beam tandem mass spectrometer that has been described in previous publications.7–11,27,28 Only a brief description is given here, emphasizing key operating parameters. For this experiment, both mass filters were operated at 2.1 MHz with a m/z range of 1–500. A sample solution of CysH+ was prepared in HPLC grade methanol–water (1 : 2 vol. ratio) containing 0.5 mM L-cysteine hydrochloride (Z99.0%, Fluka), and that of [Cys-H] was prepared in methanol–water (4 : 1) containing a mixture of 0.5 mM L-cysteine (100.1% by titration, EMD Chemicals) and sodium hydroxide (reagent grade, Fisher). The sample solution was sprayed into an ambient atmosphere through an electrospray needle at a flow rate of 0.03–0.04 mL h1. The electrospray needle was biased at 2000 to 2500 V and 2000 to 2200 V for producing positively and negatively charged species, respectively. Charged droplets entered an ion source chamber through a desolvation capillary. The capillary was held at 90 to 120 V for positive ions and 90 to 100 V for negative ions. Liquid droplets underwent desolvation as they passed through the heated capillary, converting to gas-phase ions in the source chamber. Under mild heating conditions, not all of the solvent was removed, resulting in hydrated ions.29 In the experiment, the capillary was heated to 130–160 1C for generating monohydrated ions and 110–130 1C for dihydrated ions. A skimmer with an orifice of 0.99 mm is located 3 mm from the capillary end, separating the source chamber and a radiofrequency (rf) hexapole ion guide. The skimmer was biased at 15 to 20 V for positive ions and 20 V for negative ions. Ions emerging from the skimmer were transported into the hexapole ion guide at a pressure of 20 mTorr and underwent collisional cooling and focusing.30,31 As characterized by a collision-induced dissociation (CID) experiment,7 internal energy of primary ions could be described by a Maxwell–Boltzmann distribution at B310 K. Ions subsequently passed into a quadrupole mass filter for selection of reactant ions. Reactant ions were collected and focused into an octopole ion guide. The octopole passes through a scattering cell containing neutral reactant gas. The cell pressure was measured by a Baratron capacitance manometer (MKS 690 head and 670 signal conditioner). The ion guide minimizes loss of the reactant and product ions resulting from scattering. After passing through the scattering cell, unreacted ions and product ions drifted to the end of the octopole, mass analyzed by a second quadrupole mass filter, and counted.


20497

View Article Online


PCCP

Paper

Initial kinetic energy distribution of the reactant ion beam was determined using a retarding potential analysis,32 i.e., measuring the intensity of the ion beam while scanning the DC bias voltage applied to the octopole. The DC bias voltage allowed control of the kinetic energy (ELab) of reactant ions in the laboratory frame. ELab was converted into the collision energy (Ecol) between the reactant ion and 1O2 in the centerof-mass frame using Ecol = ELab mneutral/(mion + mneutral), where mneutral and mion are the masses of 1O2 and the reactant ion, respectively. The intensity of the reactant ion beam was 2 105 counts per second and constant within 10%. The initial kinetic energy of the ion beam was 0.9 to 1.0 eV, and the energy spread was B0.6 eV which corresponds to an energy spread of B0.1 eV in the center-of-mass frame for the collisions of CysH+(H2O)1,2/[Cys-H](H2O)1,2 with 1O2. Reaction cross sections were calculated from the ratio of product and reactant ion intensities (under single ion-molecule collision conditions), the calibrated 1O2 pressure, and the effective length of the scattering cell.33 1 O2 was generated by the reaction of H2O2 + Cl2 + 2KOH O2(X3Sg and a1Dg) + 2KCl + 2H2O. We adopted this technique from Viggiano’s group34 with some modifications.8,9 Briefly, 20 mL of 35 wt% H2O2 (Acros Organics) was mixed with 13 mL of 8 M KOH (>85%, Fisher) solution in a sparger that was immersed in a cold bath maintained at 19 1C, and the resulting solution was degassed. A continuous flow of He (research grade, T. W. Smith) was then introduced to the slushy H2O2–KOH mixture at a flow rate of 50 sccm to prevent freezing of the mixture. Finally, Cl2 (Z99.5%, Sigma-Aldrich), at a flow rate of 2–3 sccm, was mixed with He in a gas proportioner and bubbled through the H2O2–KOH solution. Cl2 completely reacted with H2O2 to form ground-state and excited O2.35 The resulting gas mixture passed through a cold trap kept at 70 1C to remove water vapor. Only 3O2, O2(a1Dg) and He remained in the downstream gas. Before leaking into the scattering cell, the gases flowed through an emission cell for detection of 1O2 emission (a1Dg - X3Sg, n = 0–0) at 1270 nm.36 The emission cell was continuously pumped through a pressure control valve up to 15 Torr, so as to reduce the residence time and hence the wall quenching of 1O2 inside the cold trap, tubing and emission cell. Emission from the cell was collected by a plano-convex lens, and passed through an optical chopper (SRS model SR540) and a 0.5 nm bandwidth interference filter centered at 1270 nm. Chopped emission was focused into a thermoelectrically cooled InGaAs detector (Newport 71887 detector and 77055 TE-cooler controller), and the signal was processed by a lock-in amplifier (SRS model SR830). To determine the absolute 1O2 concentration, the detection system was calibrated using the known reaction rate for HS + 1O2 - SO + OH.10 1O2 pressure in the scattering cell is the product of the total gas pressure in the scattering cell, the percentage of Cl2 in the Cl2/He flow, and the 1O2 concentration (B10%) in the oxygen product. Collision cross section (scollision), taken as the greater of ioninduced dipole capture cross section (scapture)37 and hard-sphere collision cross section (shard-sphere), is 73–95 Å2 for CysH+(H2O)/ [Cys-H](H2O) + O2, and 82–95 Å2 for CysH+(H2O)2/[Cys-H](H2O)2 + O2 in the Ecol range of 0.05–1.0 eV. shard-sphere was estimated from

20498


the orientation-averaged contact radii of reactant ions and O2, and exceeds scapture at Ecol Z 0.1 eV. The pressure of O2/He in the scattering cell was maintained at 0.25 mTorr, which contained 5% of O2 (including 3O2 and 1O2). This pressure was chosen to provide reasonable intensities of product ions, while keeping multiplecollision effects to a minimum level. At this pressure, the probability of CysH+(H2O)1,2/[Cys-H](H2O)1,2 undergoing a single-collision with O2 is o2%, and that of double collisions is o0.1%. Note that reactant ions also collided with He, with a single-collision probability of 16–20%, and a double-collision probability of 3–7%. However, the heavy ion–light neutral combination makes these collisions insignificant compared to those with O2. Because the signals we measured are small, it is important to minimize systematic variations in experimental conditions that might be caused by drifting potentials, changes in ion beam intensities and 1O2 yield, etc. The concentration of 1O2 was monitored continuously during the experiment, and the concentration variation (controlled to be within 20%) was corrected for while calculating reaction cross sections. The entire experiment was repeated several times and each time we cycled through different Ecol values. The results presented are averages of several complete datasets. Based on the reproducibility of the cross section measurements taken over a two-month period, we estimate that the relative error is o20%. To check the reactivity of CysH+(H2O)1–2 or [Cys-H](H2O)1–2 toward 3O2 and He, control experiments were performed under the same conditions except that Cl2 was replaced by oxygen gas at the same flow rate. 2.2

Computational methods

To interpret reaction coordinates, density functional theory (DFT) electronic structure calculations were performed at the B3LYP level of theory with various basis sets including 6-31+G(d), 6-311++G(d,p) and aug-cc-pVQZ, using Gaussian 09.38 Geometries were optimized by calculating force constants at every step. All of the transition states (TSs) found were verified to be first-order saddle points, and the vibrational mode with an imaginary frequency corresponds to the associated reaction pathway. Zero-point energies (ZPEs) were scaled by a factor of 0.98139 in calculating energetics. The molecular structures of reactants, complexes, TSs and products are available on request to the corresponding author. Quasi-classical, direct dynamics trajectory simulations were used to identify important steps along the reaction coordinate for CysH+(H2O) + 1O2. Direct dynamics simulations combine electronic structure theory and molecular dynamics directly ¨dinger equation at each by solving time-independent Schro integration step, so that trajectories are calculated on the fly without the need for an analytical potential energy surface.40 In our simulations, VENUS9941 was used to set up initial conditions, and the Hessian-based predictor-corrector algorithm42 available in Gaussian 09 was used to propagate each trajectory with the Hessian matrix updated every five steps. From an analysis of the computer time required to calculate a trajectory at different levels of theory and the overall level of agreement of these different levels with experimental and/or high level benchmark results, the B3LYP/4-31G(d) level was chosen for trajectory simulations.

This journal is

c


View Article Online


Paper

PCCP

Trajectory integration was performed with a step size of 0.25 amu1/2Bohr (corresponding to a step size of B0.6 fs in trajectory time), which conserved total energy to better than 104 Hartree. The SCF = XQC option was adopted during trajectory integration so that a quadratically convergent SCF method38,43 was used in case the usual, but much faster, first-order SCF method failed to converge within the allotted number of cycles. The purpose of our trajectory simulations is to probe the gross features of the collisions between CysH+(H2O) and 1O2. Particularly, simulations were used to help discover reaction mechanisms. Thus all trajectories were calculated at an impact parameter b of 0.1 Å, rather than sampling the b distribution. Batches of trajectories were calculated at Ecol of 0.1 and 0.2 eV, respectively. Initial conditions of the reactants were chosen to mimic our experiment. Because CysH+(H2O) ions were thermalized in the experiment, their initial vibrational and rotational energies in trajectories were sampled from Boltzmann distributions at 300 K. Similarly, 300 K was used for simulating rotational and vibrational energy distributions of 1 O2. Quasi-classical initial vibrational state was simulated by giving each reactant atom displacement from equilibrium and momentum appropriate to the initial rovibrational state, with random phases for different modes. Both CysH+(H2O) and 1O2 have ZPEs in all vibrational modes. Randomly oriented CysH+(H2O) and 1O2 were given relative velocities corresponding to the simulated Ecol. All trajectories started with a center-of-mass reactant separation of 8.0 Å, and were terminated after 5000 steps or when the product separation exceeded 8.0 Å. Trajectories were calculated on an Intel core i7 6-core (3.2 GHz)-based Linux workstation cluster, and the actual computer processing time for a trajectory ranged from 200 to 350 CPU hours. gOpenMol44 was used for trajectory visualization.

III. Results and discussion 3.1

Structures of CysH+(H2O)1,2 and [Cys-H](H2O)1,2

In our previous study of CysH+ + 1O2 and [Cys-H] + 1O2, we have located the global minima for CysH+ and [Cys-H] in their conformal landscapes.9,11 Starting geometries of CysH+(H2O)1,2 and [Cys-H](H2O)1,2 were obtained by adding water to all possible hydration sites in the lowest energy conformations of CysH+ and [Cys-H], respectively, and then optimized at B3LYP/6-311++G(d,p). A similar approach was used to build the hydration shell of other amino acids.45,46 Four lowest energy conformations were found for CysH+(H2O), as shown in Fig. 1. Their structures are distinguished from each other by appending letter to structures alphabetically, e.g., CysH+(H2O)_a is the most stable conformation for CysH+(H2O). Hydration energy, indicated in parentheses below each structure, was calculated by Ehydration = E(bare ion) + nE(H2O) E(cluster), where E(bare ion), E(H2O) and E(cluster) are the B3LYP/6-311++G(d,p) energies of dehydrated ion, water and hydrated cluster, respectively; and n is the number of water ligands. As depicted in Fig. 1, both the carboxyl and ammonium groups of CysH+ offer binding sites for water. The water oxygen either binds to the –OH site of the carboxyl group with a H2O HOOC distance of 1.59 Å, or to one of the three –NH sites of the ammonium group with a H2O HN distance of 1.67–1.74 Å. The hydration energies of three –NH sites differ by a

This journal is

c


maximum of 0.14 eV, and the hydration energy of the –OH site is only 0.02 eV less than that of the strongest –NH site. Structures CysH+(H2O)2_a–f in Fig. 1 represent addition of the second water molecule to each of the four CysH+(H2O) conformers described above. The most stable dihydrated conformer, CysH+(H2O)2_a, has two water molecules hydrogen bound to –NH and –OH, respectively. Note that the two H2O molecules in CysH+(H2O)2_a occupy the hydration sites of CysH+(H2O)_a and b simultaneously, but with the total hydration energy (1.33 eV) 0.05 eV less than the sum of CysH+(H2O)_a (0.70 eV) and CysH+(H2O)_b (0.68 eV). This phenomenon, also seen in other conformers of CysH+(H2O)2, is due to decreasing effective charge on NH3+ and increasing repulsion between H2O ligands.26,47–49 Before describing the structures of [Cys-H](H2O)1,2, it is necessary to address the fact that Cys has two deprotonation sites, the carboxyl and thiol groups. pKa of the carboxyl group (2.0) is 6.2 units smaller than that of the thiol group (8.2),50 indicating that [Cys-H] should be a carboxylate in solution. However, disputes arose regarding the structure of [Cys-H] in the gas phase. Woo et al. claimed that ESI of Cys in methanol–water produced a thiolate in the gas phase according to photoelectron spectroscopy measurements.51 Tian et al. also found that the preferred deprotonation site for gaseous Cys is the thiol side chain based on H/D exchange of [Cys-H] with deuterated alcohols.52 On the other hand, Oomens et al.53 reported the gas-phase infrared multiple photon dissociation (IRMPD) spectrum of [Cys-H] which was generated by ESI of a mixture of Cys and NaOH in methanol– water. The conformation-specific IRMPD spectrum identified a carboxylate structure only for gaseous [Cys-H]. It therefore seems that the structure of [Cys-H] present in the gas phase depends on experimental conditions.53 A similar finding was reported for the carboxylate vs. phenoxide structures of gaseous deprotonated Tyr.10,53–55 In our CID experiments of [Cys-H] and [Cys-H](H2O)1,2 with Ar and 3O2, dominant product ions include HS at m/z 33. HS is a characteristic fragment of the carboxylate structure, as reported in IRMPD.53 This evidence leads us to conclude that our ESI source produced dominantly carboxylate structures for [Cys-H] and [Cys-H](H2O)1,2. For clarity, HSCH2CH(NH2)CO2 was used to denote [Cys-H] in the following discussion. Four conformers were identified for each of HSCH2CH(NH2)CO2(H2O) and HSCH2CH(NH2)CO2(H2O)2. As shown in Fig. 1, their structures can be distinguished by hydrogen bonding motifs. The most stable monohydrated conformer, HSCH2CH(NH2)CO2(H2O)_a, undergoes bidentate complexation, forming two hydrogen bonds between H2O and the carboxylate group in a six-membered cyclic arrangement,45,46,56,57 with a hydration energy of 0.65 eV. Other three monohydrated conformers are mono-dentate complexes, each forming a hydrogen bond between H2O and the carboxylate or amino group, with their hydration energies 0.09–0.21 eV less than that of HSCH2CH(NH2)CO2(H2O)_a. In conformers HSCH2CH(NH2)CO2(H2O)_c and d, the water molecule bridges –COO and –NH2, and–NH2 and –SH, respectively. Shared proton binding motifs also occur between –NH2 and –COO in HSCH2CH(NH2)CO2(H2O)_a, b and d, and between the N atom and –SH in HSCH2CH(NH2)CO2(H2O)_a, b and c. No conformer was found with the water oxygen atom


20499

View Article Online


PCCP

Paper

Fig. 1 Conformations of CysH+(H2O)1,2 and HSCH2CH(NH2)CO2(H2O)1,2 calculated at B3LYP/6-311++G(d,p). Dashed lines indicate hydrogen bonds. Bond distances are in Å. Hydration energies at 0 K (eV, including ZPE) are indicated in parentheses.

strongly bonded to an amino hydrogen atom such as in CysH+(H2O)_a. HSCH2CH(NH2)CO2(H2O)2_a represents the lowest energy conformer for dihydrated deprotonated Cys, in which two water molecules form two hydrogen bonds to carboxylate oxygen atoms, plus a weaker one between two waters. In HSCH2CH(NH2)CO2(H2O)2_b and c, one water simultaneously binds to both oxygen atoms of –COO in a manner similar to that in HSCH2CH(NH2)CO2(H2O)_a, and the other water bridges the C- and N-terminus, or the –NH2 and –SH. HSCH2CH(NH2)CO2(H2O)2_d has both water molecules inserted between the C- and N-terminus. Since this conformer is not able to include –SH in an intramolecular hydrogen bond, it has the lowest hydration energy. Our calculations indicate that protonated and deprotonated Cys have distinctly different hydration structures. Protonated Cys has higher hydration energy than the deprotonated one, presumably due to the higher positive charge density centered on –NH3+ than the diffusive negative charge density on

20500


–CO2.57,58 Similar to CysH+, HSCH2CH(NH2)CO2 has higher binding energy with the first water, while the second water is less strongly bound by 18–38% due to charge delocalization over the water molecules. 3.2

Reaction cross sections

A. CysH+(H2O)1,2 + 1O2. For the reaction of CysH+(H2O) (m/z 140) with 1O2, product ions were observed at m/z 74, 93, 107, 122, 131 and 154 over the Ecol range of 0.05–1.0 eV. Product ions of m/z 93, 107 and 122 correspond to elimination of CH2SH, HS and H2O from CysH+(H2O), respectively, of which dissociation of CysH+(H2O) to CysH+ and H2O is the most significant CID channel. Product ions of m/z 131, with an insignificant intensity, can be attributed to the formation of (CysH+)2H2O. These products were also observed upon collisions of CysH+(H2O) with 3O2/He, and therefore could be excluded from 1O2-specific reactions. Product ions of m/z 154, on the other hand, were not observed with 3O2/He. m/z 154 corresponds to

This journal is

c


View Article Online


Paper

PCCP

Fig. 2 Product cross sections for the reactions of CysH+(H2O)1,2 with 1O2, as a function of center-of-mass Ecol. Reaction efficiencies are shown on the right axis.

the formation of a hydroperoxide Cys-SOOH+, and its cross section is shown in the top frame of Fig. 2 as a function of Ecol. Also shown in Fig. 2 is the reaction efficiency (right-hand scale), calculated as sreaction/scollision. At low collision energies, CysH+ formed from CID underwent secondary reaction with 1O2, forming product ions of H2NCHCO2H+ at m/z 74.9 For the reaction of CysH+(H2O)2 (m/z 158) + 1O2, product ions were observed at m/z 74, 93, 107, 122, 131, 140, 154 and 172. Products ions of m/z 74, 93, 107, 122, and 131 have the same origins as those observed in the reaction of CysH+(H2O) + 1 O2, product ions of m/z 140 can be attributed to elimination of a H2O molecule from CysH+(H2O)2. Only m/z values of 154 and 172 belong to 1O2-specific hydroperoxide products, of which m/z 172 corresponds to a monohydrated hydroperoxide Cys-SOOH+(H2O) albeit its intensity is significantly lower than that of m/z 154 (i.e., Cys-SOOH+). Their cross sections and total reaction efficiency (sum of m/z 154 and 172) are shown in the bottom frame of Fig. 2 as a function of Ecol. Reactions of both CysH+(H2O) and CysH+(H2O)2 are significant only at low energies. Reaction efficiencies for CysH+(H2O) and CysH+(H2O)2 are 30% and 60%, respectively, at Ecol = 0.05 eV, dropping to 6% and 7% at Ecol = 0.2 eV, and becoming negligible at Ecol Z 0.5 eV. The rapid drop-off in the reaction efficiencies shown in Fig. 2 (and Fig. 3) reflects the inefficiency of the initial capture of the reactants before the energy can become dispersed among the internal modes of intermediates. Such collision energy

This journal is

c


Fig. 3 Product cross sections for the reactions of HSCH2CH(NH2)CO2(H2O)1,2 with 1O2, as a function of center-of-mass Ecol. Reaction efficiencies are shown on the right axis.

dependence suggests that reactions are complex-mediated, with complex formation probabilities and/or lifetimes that are strongly suppressed by Ecol. Another interesting observation is that the reaction of CysH+(H2O)2 is more efficient than that of CysH+(H2O), particularly at low Ecol. B. HSCH2CH(NH2)CO2(H2O)1,2 + 1O2. 1O2-specific product ions were observed at m/z 152 for HSCH2CH(NH2)CO2(H2O) (m/z 138) + 1O2, and at m/z 152 and 170 for HSCH2CH(NH2)CO2(H2O)2 (m/z 156) + 1O2. Their reaction cross sections and efficiencies as well as Ecol dependence are shown in Fig. 3. Similar to CysH+(H2O)1,2 + 1 O2, reactions of HSCH2CH(NH2)CO2(H2O)1,2 + 1O2 are strongly inhibited by collision energies (note that a similar Ecol dependence of reaction efficiency was observed for the reactions of 1O2 with dehydrated CysH+ and HSCH2CH(NH2)CO2 (ref. 9 and 11)). However, reaction efficiencies of HSCH2CH(NH2)CO2(H2O) and HSCH2CH(NH2)CO2(H2O)2 are only one-fifth and one-tenth, respectively, compared to their protonated counterparts. In addition to 1O2-specific product ions, we have observed CID product ions of HSCH2CH(NH2)CO2(H2O) and HSCH2CH(NH2)CO2(H2O)2 at m/z 33 (HS), 76 (elimination of CO2 + H2O from HSCH2CH(NH2)CO2(H2O)), 120 and 138 (elimination of a H2O molecule from HSCH2CH(NH2)CO2(H2O) and HSCH2CH(NH2)CO2(H2O)2, respectively), and 102 (elimination of H2O from the primary fragment HSCH2CH(NH2)CO2), as well as formation of the dimer (HSCH2CH(NH2)CO2)2H2O


20501

View Article Online


PCCP

Paper

at m/z 129 (insignificant intensity). In contrast to product ions of m/z 152 and 170, these product ions were observed in the control experiment with 3O2/He, and their intensities increase at high collision energies. Since these products are not relevant to 1O2 chemistry, they are not discussed further. Note that the electron detachment energy for HSCH2CH(NH2)CO2(H2O)/HSCH2CH(NH2)CO2(H2O)2 is 2.92/3.26 eV, calculated at B3LYP/6-311++G(d,p), which is beyond the Ecol range of 0.05–1.0 eV used in our experiment. Consequently, collisioninduced detachment of an excess electron from HSCH2CH(NH2)CO2(H2O)/HSCH2CH(NH2)CO2(H2O)2 could be disregarded. In addition, assuming all of the excitation energy (0.98 eV)36 and electron affinity (0.45 eV)59 of 1O2 can be used for reactions, electron transfer between HSCH2CH(NH2)CO2(H2O)/HSCH2CH(NH2)CO2(H2O)2 and 1O2 is endothermic by 1.49/1.83 eV, and thus cannot occur in our Ecol range, either. 3.3

Reaction mechanisms

A. CysH+(H2O)1,2 + 1O2. Among the conformers of CysH+(H2O) presented in Fig. 1, CysH+(H2O)_a and b are predicted to have a population of 62% and 29%, respectively, while CysH+(H2O)_c and d have a total population of 9% under our experimental conditions. Among the CysH+(H2O)2 conformers, CysH+(H2O)2_a has the largest population (>85%), while the rest of the conformers account for less than 15%. Based on the significance of their populations, we chose conformers CysH+(H2O)_a and b as the reactant structures for monohydrated CysH+, and CysH+(H2O)2_a for dihydrated CysH+, respectively, for construction of their reaction coordinates. To differentiate the water binding sites in CysH+(H2O)_a and b, we include the termini to which water binds in the formulas, i.e., CysH+(H2O)_a is referred to as CysH+(N-H2O), and CysH+(H2O)_b as CysH+(C-H2O), for clarity, in the discussion of reaction mechanisms. In a previous study, we have proposed a mechanism for the reaction of bare CysH+ with 1O2, which involves formation of a hydroperoxide intermediate Cys-SOOH+.9 Considering the similarities between the chemistry of CysH+ and its hydrated analogues, we may reasonably presume that m/z 154 and 172 correspond to formation of similar hydroperoxides from CysH+(H2O)/ CysH+(H2O)2 + 1O2. Their structures are shown in Fig. 4, and reaction enthalpies calculated at B3LYP/6-31+G(d) are as follows: CysH+(N-H2O) + 1O2 - Cys-SOOH+ + H2O

DH = 1.27 eV (1a)

CysH+(C-H2O) + 1O2 - Cys-SOOH+ + H2O DH = 1.30 eV (1b) CysH+(H2O)2 + 1O2 - Cys-SOOH+(C-H2O) + H2O DH = 1.30 eV (2a) - Cys-SOOH+(N-H2O) + H2O

DH = 1.32 eV (2b)

- Cys-SOOH (S-H2O) + H2O

DH = 1.11 eV

+

- Cys-SOOH+ + 2H2O

DH = 0.63 eV

(2c) (2d)

Fig. 4 presents potential energy surfaces (PESs) associated with possible low energy reaction pathways for CysH+(N-H2O) + 1O2

20502


and CysH+(C-H2O) + 1O2, respectively, with the reactants shown at zero energy. Energies of complexes, TSs and products are derived from B3LYP/6-31+G(d) calculations, except for that of TS2(N-H2O) which was calculated at B3LYP/aug-cc-pVQZ and compared to the energy of the reactants at the same basis set. Note that due to the mixing of open- and closed-shell characteristics of 1 O2, the standard DFT method overestimates the electronic excitation energy of 1O2.60 In our calculations, the DFT energy of 1O2 was obtained by adding the experimental excitation energy of 0.98 eV (ref. 36) to the DFT energy of 3O2. We first focus on the reaction coordinate for CysH+(N-H2O) + 1 O2, as displayed in Fig. 4a. One weakly bound complex PC(N-H2O), and several covalently bound complexes, i.e., CysOOH+(N-H2O), CysOOH+(N-H2O)0 , CysOOH+(S-H2O), Cys-SOOH+(N-H2O) and CysSOOH+(S-H2O), were identified. The transition states (TSs) connecting the complexes to each other and to the products are indicated in the figure. PC(N-H2O) can be characterized as a reactant-like complex, formed by electrostatic interaction and hydrogen bonding between CysH+(N-H2O) and O2. It has the O2 moiety sandwiched between the thiol and ammonium groups of CysH+, with distances of 2.39 and 1.71 Å for HS O–O and O–O HN, respectively. The binding energy of PC(N-H2O) is 0.05 eV with respect to the reactants. Because no rearrangement is needed to form this reactant-like complex from the reactants, it is less likely to have a significant barrier inhibiting formation of PC(N-H2O). This was verified by trajectory simulations to be discussed below. Because of a lack of directional covalent bonds between CysH+(N-H2O) and O2 moieties, PC(N-H2O) does not have a well-defined geometry at the energies available in our experiment, and is rather floppy with a large amplitude of intermolecular motion as seen in trajectories (vide infra). The point is that PC(N-H2O) allows repeated encounters between reactants, increasing reaction probability for collisions not initially in the correct geometry. To this extent, complex PC(N-H2O) acts as a ‘‘precursor complex’’. The precursor complex may undergo intramolecular proton transfer from –NH3 to the peroxide group through TS1(N-H2O), leading to formation of a persulfoxide CysOOH+(N-H2O). CysOOH+(N-H2O) is a covalently bound intermediate, with a binding energy of 0.36 eV relative to the reactants. CysOOH+(N-H2O) may then converge to a more stable hydroperoxide intermediate Cys-SOOH+(N-H2O) via TS2(N-H2O). During the interconversion from CysOOH+(N-H2O) to Cys-SOOH+(N-H2O), the proton initially abstracted by the –SOO group returns to the –NH2 group; simultaneously a H atom is transferred from the –SH group to the oxygen end of –SOO. The overall reaction pathway is outlined by black lines in Fig. 4a, i.e. CysH+(N-H2O) + 1 O2 - PC(N-H2O) - TS1(N-H2O) - CysOOH+(N-H2O) TS2(N-H2O) - Cys-SOOH+(N-H2O) - Cys-SOOH+ (m/z 154) + H2O, and referred to as pathway a. The reaction enthalpy for pathway a is 1.27 eV, and all associated TSs are below the reactants, consistent with our experimental observation of an exothermic reaction. In principle, product ions of m/z 154 could be CysOOH+ and/or CysOOH+ 0 (indicated by gray lines on the top right of Fig. 4a), which may be produced by elimination of water from CysOOH+(N-H2O). However, the reaction enthalpy for CysOOH+ (0.14 eV) as well as the barrier at TS3 (0.47 eV)

This journal is

c


View Article Online

PCCP


Paper

Fig. 4 Schematic reaction coordinate for 1O2 with (a) CysH+(N-H2O) and (b) CysH+(C-H2O). Structures of CysOOH+, CysOOH+ 0 , TS3 and TS4 in (b) are identical to those in (a). Energies of complexes, TSs, and products, relative to reactants, are derived from B3LYP/6-31+G(d) and B3LYP/aug-cc-pVQZ results. Bond distances are in Å. For TSs, vibrational modes corresponding to imaginary frequencies are indicated by displacement vectors.

leading to CysOOH+ 0 render formation of these ion structures improbable at low Ecol. Note that in pathway a, the water ligand acts as a spectator until its elimination at the last step. There exists an alternative pathway b in which the water ligand catalyzes the reaction. Pathway b was revealed by trajectory simulations where the motion of molecules is followed, allowing the molecules to show what the preferred reaction pathways are. This pathway is outlined by blue lines in Fig. 4a, corresponding to CysH+(N-H2O) + 1O2 PC(N-H2O) - TS1(N-H2O)0 - CysOOH+(N-H2O)0 - CysOOH+(S-H2O) - TS2(S-H2O) - Cys-SOOH+(S-H2O) - Cys-SOOH+ + H2O. One unique feature of pathway b is that water is intimately involved in the reaction. At TS1(N-H2O)0 , the water accepts a proton from –NH3 and in the meantime donates another to –SOO, leading to formation of CysOOH+(N-H2O)0 . CysOOH+(S-H2O), where the ‘‘S’’ denotes the side chain to which water binds, is an analogue of

This journal is

c


CysOOH+(N-H2O)0 except that the water ligand swings from –NH2 to –SOOH. The activation barrier (not included in Fig. 4a) for the interconversion between these two peroxides is minuscule (only 0.07 eV above CysOOH+(S-H2O)). The energy of CysOOH+(S-H2O) is slightly higher than that of CysOOH+(N-H2O)0 , presumably because of the absence of a –NH2 water hydrogen bond in CysOOH+(S-H2O). CysOOH+(S-H2O) may transfer a H atom from S to –NH2 via TS2(S-H2O), yielding Cys-SOOH+(S-H2O) followed by dissociation to Cys-SOOH+ + H2O. While pathways a and b yield identical products, intramolecular proton transfer in pathway a vs. waterassisted proton transfer in pathway b leads to different PESs. Fig. 4b shows the PES for CysH+(C-H2O) + 1O2. This reaction follows a similar route to pathway a for CysH+(N-H2O) + 1O2, i.e., CysH+(C-H2O) + 1O2 - PC(C-H2O) - TS1(C-H2O) CysOOH+(C-H2O) - TS2(C-H2O) - Cys-SOOH+(C-H2O) Cys-SOOH+ (m/z 154) + H2O, with an overall reaction enthalpy


20503

View Article Online


PCCP

Paper

of 1.30 eV. We did not run trajectory simulations for CysH+(C-H2O). But it seems less likely to have water assist the reaction in PC(C-H2O) since the water ligand binds to a place opposite to the reaction center. Similar to CysH+(N-H2O) + 1O2, product ions of CysOOH+ and CysOOH+ 0 could be ruled out for CysH+(C-H2O) + 1O2 due to their high endothermicities. Computation results for CysH+(H2O)2 + 1O2 at B3LYP/ 6-31+G(d) level are summarized in Fig. 5. One of the minimum energy paths leading to Cys-SOOH+(C-H2O) and/or Cys-SOOH+(N-H2O) is CysH+(H2O)2 + 1O2 - PC(H2O)2 - TS1(H2O)2 CysOOH+(H2O)2 - TS2(H2O)2 - Cys-SOOH+(H2O)2 - CysSOOH+(N-H2O)/Cys-SOOH+(C-H2O) (m/z 172) + H2O, of which TS1(H2O)2 and TS2(H2O)2 are located 0.05 and 0.1 eV below the reactants, respectively. Since one water ligand in CysH+(H2O)2 is bound to a –NH site, there exists another pathway relying on water-assisted proton transfer, as pathway b for CysH+(N-H2O). This is outlined by blue lines in Fig. 5 as CysH+(H2O)2 + 1O2 PC(H2O)2 - TS1(H2O)2 0 - CysOOH+(H2O)2 0 - CysOOH+(H2O)200 -TS2(H2O)2 0 - Cys-SOOH+(H2O)2 0 - Cys-SOOH+(S-H2O)/Cys-SOOH+(C-H2O) (m/z 172) + H2O. Note that the water-assisted pathways for CysH+(N-H2O) and CysH+(H2O)2 are extremely alike in PES and energetics. As mentioned above, CysH+(H2O)2 + 1O2 has higher reaction efficiency than CysH+(H2O) + 1O2 (see Fig. 2). This observation might be attributed to two facts. First, density of vibrational states in the dihydrated system is higher than that in the monohydrated system, and so the intramolecular vibrational relaxation is faster for prompt dissociation of Cys-SOOH+(H2O)2. Second, since the second water is less strongly bound than the first one, the dihydrated cluster has a lower energy dissociation channel to eject water than the monohydrated one. Fig. 5 includes two alternative structures for m/z 172, i.e., CysOOH+(N-H2O) and CysOOH+(C-H2O). But reaction enthalpies of CysOOH+(N-H2O) and CysOOH+(C-H2O) suggest that they are

not significant at low Ecol. Finally, a large fraction of monohydrated Cys-SOOH+(N-H2O), Cys-SOOH+(C-H2O) and Cys-SOOH+(S-H2O) may undergo secondary reactions, eliminating the remaining water ligand. Consequently, Cys-SOOH+ accounts for >85% of total product ions at Ecol r 0.1 eV. B. HSCH2CH(NH2)CO2(H2O)1,2 + 1O2. Based on their relative energies, HSCH2CH(NH2)CO2(H2O)_a accounts for 97% of monohydrated while HSCH2CH(NH2)CO2(H2O)2_a accounts for 95% of dihydrated deprotonated Cys under our experimental conditions. Therefore, these two conformers were used as the reactant ion structures for construction of the reaction coordinates for HSCH2CH(NH2)CO2(H2O) and HSCH2CH(NH2)CO2(H2O)2, respectively. Their reaction enthalpies calculated at B3LYP/6-31+G(d) are as follows, HSCH2CH(NH2)CO2(H2O) + 1O2 - HOOSCH2CH(NH2)CO2 + H2O

DH = 1.24 eV (3)

HSCH2CH(NH2)CO2(H2O)2 + 1O2 - HOOSCH2CH(NH2)CO2(H2O) + H2O - HOOSCH2CH(NH2)CO2 + 2H2O

DH = 1.34 eV (4a) DH = 0.69 eV (4b)

PESs associated with possible low energy reaction pathways for O2 with HSCH2CH(NH2)CO2(H2O) and HSCH2CH(NH2)CO2(H2O)2 are presented in Fig. 6a and b, respectively. Since these two systems present very similar reaction behaviors, we describe their PESs together. Similar to CysH+(H2O)1,2, HSCH2CH(NH2)CO2(H2O)1,2 may form precursor complexes with 1O2, i.e., PC(H2O)1,2. Their binding energies are 0.67 and 0.63 eV, respectively, with regard to the corresponding reactants. PC(H2O)1,2 goes through two consecutive TSs (TS1a(H2O)1,2 and TS1b(H2O)1,2). 1

Fig. 5 Schematic reaction coordinate for CysH+(H2O)2 with 1O2 (elimination of the second water is not shown). Energies of complexes, TSs, and products, relative to reactants, are derived from a combination of B3LYP/6-31+G(d) results. Bond distances are in Å. For TSs, vibrational modes corresponding to imaginary frequencies are indicated by displacement vectors.

20504


This journal is

c


View Article Online


Paper

PCCP

Fig. 6 Schematic reaction coordinate for 1O2 with (a) HSCH2CH(NH2)CO2(H2O) and (b) HSCH2CH(NH2)CO2(H2O)2. Energies of complexes, TSs, and products, relative to reactants, are derived from B3LYP/6-31+G(d) results. Bond distances are in Å. For TSs, vibrational modes corresponding to imaginary frequencies are indicated by displacement vectors.

At TS1a(H2O)1,2, which corresponds to a barrier slightly above PC(H2O)1,2, the H atom of the thiol group is transferred to the amino group, and the O2 moiety starts binding to the S atom. The path from TS1a(H2O)1,2 leads to TS1b(H2O)1,2, in which the same H atom now moves from –NH3 to the other end of the O2 moiety, leading to formation of a hydroperoxide complex HOOSCH2CH(NH2)CO2(H2O)1,2. HOOSCH2CH(NH2)CO2(H2O) and HOOSCH2CH(NH2)CO2(H2O)2 may liberate a water molecule to yield product ions of HOOSCH2CH(NH2)CO2 (m/z 152) and HOOSCH2CH(NH2)CO2(H2O) (m/z 170), respectively. For HSCH2CH(NH2)CO2(H2O)2 + 1O2, both HOOSCH2CH(NH2)CO2 and HOOSCH2CH(NH2)CO2(H2O) were observed experimentally; however, in contrast to what was seen in the reaction of CysH+(H2O)2, oxidation of HSCH2CH(NH2)CO2(H2O)2 produced more hydrated product ions HOOSCH2CH(NH2)CO2(H2O) than dehydrated ones. 3.4 Direct dynamics trajectory simulations of CysH+(N-H2O) + 1O2 A further understanding of the collision dynamics for CysH+(N-H2O) + 1O2 was obtained by examining their trajectories

This journal is

c


at Ecol = 0.1 and 0.2 eV. One hundred trajectories were completed at each Ecol. All trajectories were calculated at b = 0.1 Å using the B3LYP/4-31G(d) level. While a few trajectories completed reactions within trajectory simulation times (3–5 ps), the majority of the trajectories form precursor complexes and become trapped in that potential energy well. The remaining trajectories belong to nonreactive collisions, i.e., fly-by without forming long-lasting complexes. Fig. 7a demonstrates a trajectory representative of nonreactive collisions. The plots show the changes in potential energy (PE) and CM distances along trajectory simulation time. The CM distances are the distances between the centers of mass of 1O2 and CysH+, and of H2O and CysH+, respectively. This trajectory represents a direct scattering, with only one turning point in the relative motion of 1 O2 vs. CysH+, i.e., there is no sign of complex-mediation; in addition, the water ligand remains bound to CysH+ during the collision as indicated by the change in r(H2O–CysH+). The time scale of the collision is somewhat arbitrary, but two numbers are relevant. The time between the start of the trajectory and the onset of strong interaction, which depends on the reactant orientation, is around 350 fs. The time taken for reactants to approach within 5 Å of CM


20505

View Article Online

Paper


PCCP

Fig. 7 Representative plots of a (a) nonreactive, (b) complex-forming, and (c) complete reactive trajectory at Ecol = 0.1 eV. Trajectory (c) illustrates water-assisted proton transfer.

distance is around 250 fs. During the trajectory, PE fluctuates due to the vibrational motions of the reactants. Fig. 7b illustrates a complex-forming trajectory, with the similar reactants approach time as that for the nonreactive trajectory. In addition to the changes in PE and the CM distances r(O2–CysH+) and r(H2O–CysH+), Fig. 7b shows the approaching of the O2 moiety toward the S atom as indicated by the distances between S and two O atoms of O2, i.e., rSO and rSO0 . This trajectory initially forms a loosely bound complex as shown by the decrease of r(O2–CysH+) to less than 5 Å starting at 450 fs. The reactants have repeated encounters after their initial collision and remain close to each other. In agreement with the PES in Fig. 4a, no obvious potential barrier is observed during formation of the precursor complex. Toward the end of the trajectory, both rSO and rSO0 decrease rapidly, implying that the trajectory may ultimately lead to formation of a persulfoxide complex. Finally, a trajectory producing Cys-SOOH+ and a separated water molecule is depicted in Fig. 7c. The bonds plotted in

20506


Fig. 7c correspond to the N–H bonds being broken and formed in –NH3, the O–H bonds being broken and formed in H2O, the breaking thiol group, and the new O–H and S–O bonds being formed in the product Cys-SOOH+, respectively. High frequency oscillations of various bonds reflect the vibrations of the reactants or products. Two middle frames of Fig. 7c show a concerted transfer of two protons at B700 fs; one proton is transferred from –NH3 to water, and simultaneously another from water to –SOO. At the same time, a persulfoxide bond is formed between the O2 moiety and S, leading to CysOOH+(N-H2O). Around 1000 fs later, the H in the thiol group is transferred to the amino group and the water swings from –NH3 to –SOOH, forming CysSOOH+(S-H2O). This trajectory verifies reaction pathway b we have proposed in Fig. 4a. Elimination of water occurs after completion of proton and hydrogen transfer. By the end of the trajectory, water is separated from Cys-SOOH+ by 8.5 Å. The simulations show that all persulfoxide and hydroperoxide complexes formed in trajectories did not decay back to

This journal is

c


View Article Online


Paper

PCCP

the reactants before the termination of the trajectories (typically 3–5 ps). This suggests that the lifetimes of these complexes are at least no less than the trajectory time. For comparison, the classical rotational period of a complex estimated using the average angular momentum is 2.5 ps at Ecol = 0.1 eV. The flyby time, taken as the time required for 5 Å motion at the relative speed of reactants, is 0.58 ps at Ecol = 0.1 eV. The fly-by time gives a measure of how long a direct collision would last at the same Ecol. Clearly, the complex lifetime is significantly longer than the fly-by time, and comparable to the complex rotational period. Note that one problem for the present system is that the combined lifetime of the precursor and peroxide complexes is too long to allow most trajectories to complete. But trajectories provide information concerning the early time dynamics; particularly, confirming complex mediation in the reaction.

3.5 Role of water molecules and its implication in mimicking solution-phase oxidation Hydroperoxide complexes Cys-SOOH+ and HOOSCH2CH(NH2)CO2 have been formed during the reactions of 1O2 with bare CysH+ and HSCH2CH(NH2)CO2,9,11 respectively, as verified by trajectory simulations.9,11 However, in those cases, nascent Cys-SOOH+ ultimately decomposes to H2NCHCO2H+ + CH3SH + 3O2 (with a dissociation energy of 0.67 eV), and HOOSCH2CH(NH2)CO2 decomposes to NH2CH2CO2 + CH2S + 3O2 (with a dissociation energy of 2.12 eV). These dissociation reactions need to be driven by the 1O2 excitation energy, the so-called dissociative excitation energy transfer.61 Dissociative excitation energy transfer breaks the spin conservation rule, and therefore requires strong spin-orbital coupling within these complexes which could catalyze access to triplet channels.9,11,61 However, the dissociative excitation energy transfer mechanism completely shuts off in hydrated ions. Instead, we have observed stable Cys-SOOH+ and HOOSCH2CH(NH2)CO2 products produced by elimination of water ligands from Cys-SOOH+(H2O)1,2 and HOOSCH2CH(NH2)CO2(H2O)1,2, respectively. The key to unraveling this puzzle lies in the fact that ion clusters with weakly bound water provide a mechanism by which the energized hydroperoxide product complex can dispose of sufficient internal excitation so that the hydroperoxide moiety does not undergo further dissociation. The dissociation energies of Cys-SOOH+(N-H2O), Cys-SOOH+(C-H2O) and CysSOOH+(S-H2O) to Cys-SOOH+ + H2O are 0.69, 0.67 and 0.49 eV, respectively; while those of Cys-SOOH+(H2O)2 to Cys-SOOH+(N-H2O) + H2O, Cys-SOOH+(C-H2O) + H2O and Cys-SOOH+(S-H2O) + H2O are 0.62, 0.64 and 0.83 eV, respectively. Similar dissociation energies were found for HOOSCH2CH(NH2)CO2(H2O)1,2. Therefore, CysSOOH+(H2O)1,2 and HOOSCH2CH(NH2)CO2(H2O)1,2 are expected to dissociate by loss of water molecule(s), and retain the structures of peroxide moieties. These dissociation reactions circumvent dynamical bottlenecks imposed by a spin-forbidden process (i.e., formation of 3O2), and should be more efficient. As a result, just one water molecule could introduce a significant step in the transition of Cys oxidation from the gas phase to the solution, and hydrated Cys ions generally have higher reaction efficiencies with 1O2 than dehydrated Cys ions.

This journal is

c


A complication in thinking about the reactions of hydrated clusters with 1O2 is that the water ligands may physically quench 1O2 during collisions. In the present experiment, we were not able to directly probe the physical quenching of 1O2. The quasi-classical trajectory method we used cannot simulate the physical quenching of 1O2, either, because trajectories were confined to the lowest energy singlet potential energy surface. However, trajectory simulations of CysH+(H2O) + 1O2 illustrate that at Ecol = 0.1 eV, only less than 18% of collisions have 1O2 attack the water ligand directly, and only a fraction of such collisions may actually quench 1O2. Therefore, it is less likely that the physical quenching by water would significantly affect the branching of 1O2 chemical reactions. An experiment using D2O hydrated clusters could decrease the rate of physical quenching (the rate constant for 1O2 quenching is 1.8 104 s1 in D2O solution vs. 2.4 105 s1 in H2O (ref. 62)), and may provide more insight into the relative physical vs. chemical quenching contributions in the collisions of hydrated Cys with 1O2. The above scheme may lead to an impression that the water in hydrated Cys clusters acts mostly as a spectator in reactions such as rare gas tagging (in rare gas-tagged clusters, rare gas atoms represent very weak perturbations to hosts, and spectra of rare gas-tagged clusters represent structures and dynamics of unperturbed systems).63–65 However, this is not the case for oxidation of hydrated Cys. Trajectories show water-catalyzed proton transfer for CysH+(N-H2O) + 1O2, demonstrating that the reaction coordinate of CysH+ + 1O2 can be altered by the absorbed water. This finding reinforces the understanding that gas-phase hydrated amino acids involve both solute and solvent dynamics and their coupling, rather than simply amino acid dynamics under the influence of some representation of the solvent.66

IV. Conclusions Guided-ion-beam tandem mass spectrometry was employed to determine reaction products, cross sections and collision energy dependence for the reactions of O2(a1Dg) with monoand dihydrated protonated/deprotonated Cys. DFT calculations were carried out to identify reaction coordinates, investigate thermodynamics and energy barriers. Quasi-classical, direct dynamics trajectory simulations were used to study the atomic-level mechanism of CysH+(N-H2O) + 1O2. The combined experimental and theoretical investigation reveals the effects of hydration on Cys oxidation by 1O2. It is found that hydrated systems are able to relax ‘‘energized, unstable’’ nascent hydroperoxide products by ‘‘evaporating’’ water ligand(s); consequently, they are able to present similar reaction behavior as observed in solution-phase photooxidation. In particular: (1) a single water molecule is sufficient to stabilize the hydroperoxide product; (2) conformers with water bound at different sites may follow different reaction pathways. For example, hydrated CysH+ with water bound to the ammonium group shows water-assisted proton transfer; (3) addition of the second water molecules significantly enhances the reaction efficiency of CysH+, but has little effect on that of deprotonated Cys;


20507

View Article Online

PCCP

Paper

(4) in spite of the resemblances between the reaction mechanisms of hydrated protonated and deprotonated Cys (as both show complex-mediation and formation of persulfoxide/hydroperoxide followed by elimination of water), hydrated CysH+ shows much higher efficiencies in forming hydroperoxides with 1O2 than hydrated deprotonated Cys, mimicking the pH dependence observed in solution-phase photooxidation of Cys.25


Acknowledgements This work was supported by the National Science Foundation CAREER Award (#CHE-0954507), Queens College Research Enhancement Funds, and PSC-CUNY Research Awards. We thank Yigang Fang for his assistance with LabVIEW programing for mass spectrometry experiments, and Alec Greer (CUNY Brooklyn) for his valuable comments on the manuscript.

References 1 M. J. Davies, Biochem. Biophys. Res. Commun., 2003, 305, 761–770. 2 M. J. Davies, Biochim. Biophys. Acta, 2005, 1703, 93–109. 3 G. Palumbo, Expert Opin. Drug Delivery, 2007, 4, 131–148. 4 K. G. McGregor and C. Anastasio, Atmos. Environ., 2001, 35, 1091–1104. 5 C. S. Foote, Science, 1968, 162, 963–970. 6 C. Schweitzer and R. Schmidt, Chem. Rev., 2003, 103, 1685–1757. 7 Y. Fang and J. Liu, J. Phys. Chem. A, 2009, 113, 11250–11261. 8 Y. Fang, F. Liu, A. Bennett, S. Ara and J. Liu, J. Phys. Chem. B, 2011, 115, 2671–2682. 9 F. Liu, Y. Fang, Y. Chen and J. Liu, J. Phys. Chem. B, 2011, 115, 9898–9909. 10 F. Liu, Y. Fang, Y. Chen and J. Liu, J. Phys. Chem. B, 2012, 116, 6369–6379. 11 Y. Fang, F. liu, R. Emre and J. Liu, J. Phys. Chem. B, 2013, 117, 2878–2887. 12 M. Yamashita and J. B. Fenn, J. Phys. Chem., 1984, 88, 4451–4459. 13 J. B. Fenn, M. Mann, C. K. Meng, S. F. Wong and C. M. Whitehouse, Science, 1989, 246, 64–71. 14 D. Gerlich, in State-Selected and State-to-State Ion-Molecule Reaction Dynamics. Part I. Experiment, ed. C. Y. Ng and M. Baer, John Wiley & Sons, Inc., New York, 1992, vol. 82, pp. 1–176. 15 D. Liu, T. Wyttenbach and M. T. Bowers, Int. J. Mass Spectrom., 2004, 236, 81–90. 16 P. H. Fishman, J. W. Kusiak and J. M. Bailey, Biochem., 1973, 12, 2540–2544. 17 G. Gennari, G. Cauzzo and G. Jori, Photochem. Photobiol., 1974, 20, 497–500. 18 S. Cannistraro, G. Jori and A. van der Vorst, Photochem. Photobiol., 1978, 27, 517–521. 19 R. C. Straight and J. D. Spikes, in Singlet O2, Vol. 4: Polymers and Biomolecules, ed. A. A. Frimer, CRC Press, Boca Raton, Florida, 1985, pp. 91–143.

20508


20 W. Ando and T. Takata, in Singlet O2,Vol. 3: Reaction Modes and Products Part 2, ed. A. A. Frimer, CRC Press, Boca Raton, 1985, pp. 1–117. 21 M. Rougee, R. V. Bensasson, E. J. Land and R. Pariente, Photochem. Photobiol., 1988, 47, 485–489. 22 G. Z. Justo, F. A. Camargo, M. Haun, A. Faljoni-Alario and N. Duran, Physiol. Chem. Phys. Med. NMR, 2000, 32, 145–154. 23 M. J. Davies, Photochem. Photobiol. Sci., 2004, 3, 17–25. 24 E. L. Clennan, Acc. Chem. Res., 2001, 34, 875–884. 25 A. A. Frimer, Singlet O2, Vol III, Reaction Modes and Products, Part 2, CRC Press, Boca Raton, FL, 1985. 26 H. Wincel, Int. J. Mass Spectrom., 2006, 251, 23–31. 27 Y. Fang, A. Bennett and J. Liu, Int. J. Mass Spectrom., 2010, 293, 12–22. 28 Y. Fang, A. Bennett and J. Liu, Phys. Chem. Chem. Phys., 2011, 13, 1466–1478. 29 S. E. Rodriguez-Cruz, J. S. Klassen and E. R. Williams, J. Am. Soc. Mass Spectrom., 1997, 8, 565–568. 30 R. M. Moision and P. B. Armentrout, J. Am. Soc. Mass Spectrom., 2007, 18, 1124–1134. 31 A. N. Krutchinsky, I. V. Chernushevich, V. L. Spicer, W. Ens and K. G. Standing, J. Am. Soc. Mass Spectrom., 1998, 9, 569–579. 32 K. M. Ervin and P. B. Armentrout, J. Chem. Phys., 1985, 83, 166–189. 33 P. B. Armentrout, J. Anal. At. Spectrom., 2004, 19, 571–580. 34 A. Midey, I. Dotan and A. A. Viggiano, J. Phys. Chem. A, 2008, 112, 3040–3045. 35 A. J. Midey, I. Dotan and A. A. Viggiano, Int. J. Mass Spectrom., 2008, 273, 7–10. 36 W. J. Lafferty, A. M. Solodov, C. L. Lugez and G. T. Fraser, Appl. Opt., 1998, 37, 2264–2270. 37 J. Troe, Chem. Phys. Lett., 1985, 122, 425–430. 38 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Rev. B. 01, Gaussian, Inc., Wallingford, CT, 2009. 39 J. Zheng, I. M. Alecu, B. J. Lynch, Y. Zhao and D. G. Truhlar, Database of Frequency Scale Factors for Electronic Model Chemistries, Version 2, http://comp.chem.umn.edu/freqs cale/version2.htm, 2010.

This journal is

c


View Article Online


Paper

PCCP

40 W. L. Hase, K. Song and M. S. Gordon, Comput. Sci. Eng., 2003, 5, 36–44. 41 W. L. Hase, K. Bolton, P. de Sainte Claire, R. J. Duchovic, X. Hu, A. Komornicki, G. Li, K. Lim, D. Lu, G. H. Peslherbe, K. Song, K. N. Swamy, S. R. Vande Linde, A. Varandas, H. Wang and R. J. Wolf, VENUS 99: A general chemical dynamics computer program, Texas Tech Univeristy Lubbock, TX, 1999. 42 V. Bakken, J. M. Millam and H. B. Schlegel, J. Chem. Phys., 1999, 111, 8773–8777. 43 G. B. Bacskay, Chem. Phys., 1981, 61, 385–404. 44 L. Laaksonen, gOpenMol, Center for Scientific Computing, Espoo, Finland, 3.0 ed., 2005, available at www.csc.fi/gopenmol/. 45 Y. Yao, D. Chen, S. Zhang, Y. Li, P. Tu, B. Liu and M. Dong, J. Phys. Chem. B, 2011, 115, 6213–6221. `te and D. Jacquemin, 46 C. Michaux, J. Wouters, E. A. Perpe J. Am. Soc. Mass Spectrom., 2009, 20, 632–638. 47 M. Rozman, J. Am. Soc. Mass Spectrom., 2005, 16, 1846–1852. 48 B. Gao, T. Wyttenbach and M. T. Bower, J. Am. Chem. Soc., 2009, 131, 4695–4701. 49 H. Wincel, Chem. Phys. Lett., 2007, 439, 157–161. 50 P. Vollhardt and N. Schore, Organic Chemistry, W. H. Freeman and Company, New York, 2009. 51 H.-K. Woo, K.-C. Lau, X.-B. Wang and L.-S. Wang, J. Phys. Chem. A, 2006, 110, 12603–12606. 52 Z. Tian, A. Pawlow, J. C. Poutsma and S. R. Kass, J. Am. Chem. Soc., 2007, 129, 5403–5407.

This journal is

c


53 J. Oomens, J. D. Steill and B. Redlich, J. Am. Chem. Soc., 2009, 131, 4310–4319. 54 Z. Tian and S. R. Kass, J. Am. Chem. Soc., 2008, 130, 10842–10843. 55 Z. Tian, X.-B. Wang, L.-S. Wang and S. R. Kass, J. Am. Chem. Soc., 2009, 131, 1174–1181. 56 D. Liu, T. Wyttenbach, C. J. Carpenter and M. T. Bowers, J. Am. Chem. Soc., 2004, 124, 3261–3270. 57 H. Wincel, J. Am. Soc. Mass Spectrom., 2008, 19, 1091–1097. 58 T. Wyttenbach, D. Liu and M. T. Bowers, Int. J. Mass Spectrom., 2005, 240, 221–232. 59 K. M. Ervin, I. Anusiewicz, P. Skurski, J. Simons and W. C. Lineberger, J. Phys. Chem. A, 2003, 107, 8521–8529. 60 J. M. Matxain, M. Ristilae, A. Strid and L. A. Eriksson, Chem.–Eur. J., 2007, 13, 4636–4642. 61 A. A. Viggiano, A. Midey, N. Eyet, V. M. Bierbaum and J. Troe, J. Chem. Phys., 2009, 131, 094303. 62 F. Wilkinson, W. P. Helman and A. B. Ross, J. Phys. Chem. Ref. Data, 1995, 24, 663–1021. 63 M. B. Knickelbein and W. B. Menezes, Phys. Rev. Lett., 1992, 69, 1046–1049. 64 W. J. C. Menezes and M. B. Knickelbein, J. Chem. Phys., 1993, 98, 1856–1866. 65 G. Gregoire and M. A. Duncan, J. Chem. Phys., 2002, 117, 2120–2130. 66 J. H. Jensen and M. S. Gordon, J. Am. Chem. Soc., 1995, 117, 8159–8170.


20509

Repeatability of kinematic and electromyographical measures during standing and trunk motion: how many trials are sufficient?

Cysteine-mediated mechanism disrupts energy transfer to prevent photooxidation.

How many work hours are requisite to publish a manuscript?

How Many Genes are Needed to Resolve Phylogenetic Incongruence?

How many cancers are attributable to occupational exposures?

How many suicide terrorists are suicidal?

How many dimensions are needed to describe pain properly?

How many scientific papers are not original?

Implementation of approach bias re-training in alcoholism-how many sessions are needed?

Exposures to conducted electrical weapons (including TASER® devices): how many and for how long are acceptable?

Consultation-liaison psychiatrists. How many are there and how are they funded?

How many is too many?

How many longitudinal covariate measurements are needed for risk prediction?

How many new cancers are diagnosed after emergency admission?

MediData: how many types of medical marijuana are for sale?

How many doctors are needed in general practice?

How many genes are selected in populations of Dacus oleae.

Notification of tuberculosis: how many cases are never reported?

Assessing the visual vertical: how many trials are required?

Assessment of maximal handgrip strength: how many attempts are needed?

A modular approach to neutral P,N-ligands: synthesis and coordination chemistry.

Why many semiempirical molecular orbital theories fail for liquid water and how to fix them.

How cyclobutanes are assembled in nature--insights from quantum chemistry.

How Many Ligands Can Be Bound by Magnesium-Porphyrin? A Symmetry-Adapted Perturbation Theory Study.